Twin photodetectors wearable PPG sensor for controlled measurement and improved signal quality index at lower sample rate

• Published in: AIP conference proceedings Vol. 2372; no. 1
• Main Authors: Hasan, Muhideen Abbas, Khudhair, Ahmed Kareem, Keream, Settar S.
• Format: Journal Article Conference Proceeding
• Language: English
• Published: Melville American Institute of Physics 11.11.2021
• Subjects: Accelerometers; Algorithms; Heart rate; Photometers; Real time; Signal distortion; Signal processing; Signal quality; Three axis; Wearable technology
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0065870

Wearable Photoplethysmograph (PPG) signal distortion due to motion artifacts (MA) induced by everyday movements was a limitation in the development of real-time health monitors. The task of removing the MA is usually an expensive computational procedure due to the complex operations performed to obtain a clean, reliable PPG signal for measuring heart rate. Obtaining an effective PPG signal quality index using a simple algorithm with a lower sample rate poses a major challenge in processing such signals. Data of corrupted PPG signals by MA were collected from fifteen healthy subjects while four kinds of different daily motions. To remove these artifacts during real-time monitoring, a least mean square (LMS) based adaptive noise cancellation method is implemented to twin photodetectors wearable PPG sensor outputs. In contrast to the traditional use of a three-axis accelerometer to provide the MA, experiments have shown that the proposed method efficiently recovers pulse from damaged PPG while improving the signal quality index (SQI) at a lower sample rate. By implementing this approach, it is possible to provide wearable PPG sensors for real-time health monitoring with lower cost and higher accuracy.